82148954 74000781 25 May 2007 3:55:16 UTC 4 Jun 2007 3:55:16 UTC In Progress Unknown New --- --- ---
82101448 73902364 24 May 2007 21:37:07 UTC 25 May 2007 8:32:58 UTC Over Success Done 21,983.36 40.16 25.24
82074316 73931201 24 May 2007 18:27:54 UTC 25 May 2007 3:55:16 UTC Over Success Done 22,391.50 40.91 25.52
81942999 73809341 24 May 2007 3:30:18 UTC 24 May 2007 18:27:54 UTC Over Success Done 20,748.55 37.91 25.75
81893125 71623280 23 May 2007 21:21:14 UTC 24 May 2007 15:21:13 UTC Over Success Done 20,795.00 37.99 25.67
81834544 73709038 23 May 2007 15:18:45 UTC 24 May 2007 3:30:18 UTC Over Success Done 20,081.49 36.89 24.65
81800874 73678184 23 May 2007 11:39:35 UTC 23 May 2007 21:21:14 UTC Over Success Done 14,023.86 25.76 34.05
81754602 73635627 23 May 2007 6:33:31 UTC 23 May 2007 15:18:45 UTC Over Success Done 13,413.26 24.64 30.03
Hello,
I was told that these BIG w/u would be receiving "More" credit due too there length.Yet again last night had another Huge w/u and only turned in 3 yesterday compared to 6 a day.Why are these w/u so large.https://boinc.bakerlab.org/rosetta/result.php?resultid=82101448
This is one of the 1 decoy w/u that take so long, and receive , less credit.I was told these 1 decoy w/u are good for Rosetta, and more credit will be received.
"Not so" I am receiveing Lesss credit for a w/u that takes twice the time? Why?

THX

Doug

---

Doug, I'm not sure how much reading you've been doing or how long you've been around these boards, but in a word, the reason why is that other people's machines found these tasks were not as great a hurdle as yours did. The raw computational work for each type of task is slightly different. Always highly concenrated on floating point operations, but slightly different.

Your credit issued is based on the credit claims of all those returning results for the same study before you. For the "short" ones, your machine found they crunched more easily then your BOINC benchmarks predicted and you were granted more credit then you claimed. For the "long" ones, other people's machine found them to crunch more easily then their BOINC benchmarks predicted, and they were granted more then they claimed.

There is also variability with a given protein or RNA being studied from model to model, and noone knows ahead of time what a given random starting point will yield. You've only got two of the shorter tasks left in your history. So it is entirely possible that these were unusually high credit for you.

...yes, looking at the rest of your history, you see that the granted credit of about 32 for 13-14,000 seconds of work is abnormally high compared to the rest of your history. ...on the other hand you have one that was granted almost 42 credits for 13,567 seconds of work as well. Another was granted 26 credits for almost exactly same number of seconds.

All of your "long" tasks were new RNA studies. And Rhiju has improved some of the energy functions being used in the new release which just came out today. And these types of improvements will help make the crunch times per model more consistent and remove much of the variability. These RNA studies are still very new science, so I'm sure there is still much to learn about hunting down the best models efficiently.

---

Rosetta Moderator: Mod.Sense

Doug, I'm not sure how much reading you've been doing or how long you've been around these boards, but in a word, the reason why is that other people's machines found these tasks were not as great a hurdle as yours did. The raw computational work for each type of task is slightly different. Always highly concenrated on floating point operations, but slightly different.

Your credit issued is based on the credit claims of all those returning results for the same study before you. For the "short" ones, your machine found they crunched more easily then your BOINC benchmarks predicted and you were granted more credit then you claimed. For the "long" ones, other people's machine found them to crunch more easily then their BOINC benchmarks predicted, and they were granted more then they claimed.

There is also variability with a given protein or RNA being studied from model to model, and noone knows ahead of time what a given random starting point will yield. You've only got two of the shorter tasks left in your history. So it is entirely possible that these were unusually high credit for you.

...yes, looking at the rest of your history, you see that the granted credit of about 32 for 13-14,000 seconds of work is abnormally high compared to the rest of your history. ...on the other hand you have one that was granted almost 42 credits for 13,567 seconds of work as well. Another was granted 26 credits for almost exactly same number of seconds.

All of your "long" tasks were new RNA studies. And Rhiju has improved some of the energy functions being used in the new release which just came out today. And these types of improvements will help make the crunch times per model more consistent and remove much of the variability. These RNA studies are still very new science, so I'm sure there is still much to learn about hunting down the best models efficiently.

Thx again MOD Sense,
I am under the suspiscian that at night the w/u are slowing down or something.Yes, I donnot do much reading at the Forums here due to time constraints.Surfing the internet is a Luxury for this 1 BOX, when trying to run Boinc 24-7 and just Rosey.One of the Most important experiments due too the great work being done, I am just guessing when it comes too decoys, but, I believe the 1 decoy w/u are good for Rosey.That is all I wanted too convey, albeit, conveying and Communicating before a morning coffee is not a good idea.
Am happy to receive a used pc that I will fix up and use for surfing,Then this Box can crunch 24-7, without interuption.I know very little about pc,s this used one is a Compact Presario 7469 with Windows 98 installed.Will add RAM and change it over to a Linux based Distro, probably Feisty Dawn from Ubuntu.This Box has PClinuxOS, very fast, clean , and FREE!

"Happy Crunching ev1"
Sincerely
Doug
Here is a real different w/u that found https://boinc.bakerlab.org/rosetta/result.php?resultid=81346627WOW
161 decoys with 161 attemts.Find it quite interesting the differences in w/u,s and
deoys found.

---

I've been doing my forementioned study of different crunch time settings. Within this are included both the large decoy/runtime and small decoy/runtime. Thus far, I've seen little difference in the amount of "granted credit/hour" for either runtime or # of decoys done. I'm presently doing the 12 hour runs, then tomorrow I switch to the 24 hour setting for several weeks. I don't see much difference across a range of machines from Celeron 500 to AMD64 X2 6000.

--

Here are two units...

https://boinc.bakerlab.org/rosetta/result.php?resultid=82358826
https://boinc.bakerlab.org/rosetta/result.php?resultid=82427417

The first unit was crunched on an AMD-X2-5200 and the second on an AMD-X2-4200
Both systems have 2GB of DDR2-667 RAM..

Unit one....

43011.1 seconds 53 decoys
Claimed credit	212.046520367728
Granted credit	163.87981395461 

Unit two....

42994.3 seconds 46 decoys
Claimed credit	126.624917217974
Granted credit	142.570557789278 

While granted credit works out to around 3.1 per decoy generated, there is a rather large disparity between the claim and the grant on the faster machine... I suspect that neither system found the unit much of a challenge... It's just a bit strange... And yes the granted credit on the faster machine was down about 25% from 'normal' work units that run for a smilar number of seconds...

---

Looking for a team ??? Join BoincSynergy!!

I've been doing my forementioned study of different crunch time settings. Within this are included both the large decoy/runtime and small decoy/runtime. Thus far, I've seen little difference in the amount of "granted credit/hour" for either runtime or # of decoys done. I'm presently doing the 12 hour runs, then tomorrow I switch to the 24 hour setting for several weeks. I don't see much difference across a range of machines from Celeron 500 to AMD64 X2 6000.

Hi Astro....great study! Thanks for the info as this is something we've wondered about as well.....Cheers, Rog.

I've been doing my forementioned study of different crunch time settings. Within this are included both the large decoy/runtime and small decoy/runtime. Thus far, I've seen little difference in the amount of "granted credit/hour" for either runtime or # of decoys done. I'm presently doing the 12 hour runs, then tomorrow I switch to the 24 hour setting for several weeks. I don't see much difference across a range of machines from Celeron 500 to AMD64 X2 6000.

Hi Astro....great study! Thanks for the info as this is something we've wondered about as well.....Cheers, Rog.

Thanks, not sure how accurate it's going to be. What I mean by that, is that the # of samples is going to be LOW (IMO). Around 50 per setting (higher for 3hour, and less for 24 hour). I'm switching to 24 in the next couple hours, so it won't be complete for almost two more months. I may get impatient and lower the number of 24 hour samples even more. Time will tell.

Just switched to 24 hour. The more I think about it the more I think an accurate sample size isn't really possible. If a small number of samples are chosen, then they wu types play a large role in the results, and that defeats the study on the run times themselves. If large sample is chosen, then it will range across many executables (applications), and might vary because of that. I don't know. I guess what I end up with is, what I end up with.

... I am receiveing Lesss credit...

As my 1.4 GHz Celeron with massive 256 KB L2 cache gently weeps. Back to sub-100.